Jamel Hammoud

Using Dreambooth (Stable Diffusion) to generate self-portraits.

These are just some of the hundreds of unedited AI-generated images I was able to generate using @mysteryguitarman's Dreambooth Stable Diffusion implementation. With "only" 5 hours of training on 2 3090s, based on 16 photos of my face and body, I have my own AI model that can generate images of myself according to any prompt I provide. It's honestly mind-blowing...